Funding opaque long-term assets with demand deposits creates rollover risk. This crisis has revealed that short term wholesale lenders are very prone to run, and that maturity mismatch accelerates fire sales and causes further liquidity calls (Brunnermeier, 2009).
Basel capital requirements focus on individual banks’ asset risk, underplaying a classic maturity transformation risk. They failed to address the temporal dynamics of funding risk. In a panic situation the issue is no longer the cost but the availability of funding. Relying on frequent rollovers amplifies the speed of fire sales and further reinforcement to panic.
Simple remedies such as high reserve requirements, or extending government insurance on wholesale short term funding, seem too expensive (in terms of cost and incentives) or obsolete.
What else can be done?
One group of economists – Brunnermeier et al. (2009) – propose higher capital requirements indexed to:
- asset growth,
- total leverage, and
- maturity mismatch.
We favor a related approach, namely ‘liquidity and capital insurance’, which we believe offers better incentives and some advantages in terms of likely political support.
The case for liquidity charges
There is consensus that some elements of future regulation should work like Pigovian taxes, discouraging strategies that create systemic risk. Our proposal is to establish a mandatory liquidity charge, to be paid continuously to a regulator who is able to provide emergency liquidity (and perhaps capital) during systemic crisis. The charge should be increasing in the maturity mismatch of assets and liabilities, and would be applicable to all institutions with access to safety net guarantees. Its effect should be to make short and medium term (up to one year) bank funding comparable in cost. Retail deposits would be exempted, as they are more stable thanks to their own separate insurance.
Revenues would go into an Emergency Liquidity Insurance Fund (ELIF), with legal autonomy and pre-packaged access to central bank liquidity and government funds backing. Upon significant aggregate liquidity runs (not concerning single banks), the payment of insurance would be triggered by the relevant supervisor, resulting in immediate liquidity support, guarantees on uninsured wholesale funding, and some automatic capital injections. Specific conditions may be attached, such as restrictions on compensation and dividends, as well as on some strategic choices.
Insurance charges as prepayment for future rescue costs
The insurance charges could be thought of as prepayment for future rescue costs. Yet the main goal is not to make banks pay upfront for future damage, but to realign funding incentives among beneficiaries of the safety net. Reducing reliance on short term market funding would reduce the spreading of panic in a confidence crisis, and ultimately systemic risk.
Deposit withdrawal risk is natural for banks intermediating between retail preferences for immediacy and long term funding of production. Yet the short term bias in wholesale funding is not equally justified.1 The lower cost of short term funding reflects the fact that short term lenders bear little risk, which is shifted to other stakeholders, such as capital and taxpayers. Thus the charges would make banks properly internalize the potential damage caused to others.
Liquidity insurance versus capital requirements
The charge for liquidity risk could be seen as a liquidity insurance premium – a pre-payment for the contingent support that banks eventually receive during those episodes. As such, it can make emergency intervention politically more acceptable, especially after the current bail-outs.
We believe that higher bank capital ratios alone are not the ideal solution to the issue of bank liquidity risk.
To prevent a systemic crisis, banks’ own capital would need to be very large during normal times. This has several disadvantages. Shareholders may be tempted to see bank capital as an asset to which they are fully entitled. Banks with plenty of capital on their books may try to “lever it up”, not necessarily through leverage (which is constrained by capital requirements) but through riskier investment strategies. Shareholders’ claims on bank capital also create delicate legal issues in case of supervisory intervention, since seizing a bank ahead of a formal default may be seen as a violation of private property rights, and may lead to lawsuits. This has been cited as an impediment to earlier or more drastic interventions in recent cases.
In contrast, our liquidity insurance scheme would arrange for a contingent injection of capital in systemic crises only, like in the capital insurance scheme proposed by Kashyap, Rajan, and Stein (2008). Our approach relies on a public regulator and is cheaper, as long as most support provided is temporary, based on central bank liquidity and loan guarantees (Suarez, 2008).
Do liquidity charges penalize systemic risk creation?
Liquidity charges should reflect the marginal contribution of banks’ short term funding decisions to the generation of systemic costs. Systemic risk represents the simultaneous realization of tail risks, which are hard to estimate. Extreme loss co-movements are rarely observed, and may be triggered by a different asset class each time. However, liquidity runs are present in the escalating phase of all systemic crises and have a clearly negative amplifying effect. Thus we propose short term wholesale liabilities, properly weighted by the bank’s maturity mismatch, be taken as a proxy of the bank’s marginal contribution to systemic risk.
This liquidity charge should be levied at high frequency (say, weekly). It should reflect the scale of uninsured short term funding, the degree of maturity mismatch and, ideally, the term premia at the short end of the yield curve (say, up to one year, in order to correct term-preference distortions associated with the slope of the yield curve along the business cycle).
Side effects and institutional details
Arguably, a liquidity charge on regulated banks could shift short term funding to a shadow banking sector. But this would not be a problem if shadow banks enjoyed only limited recourse to the regulated banks. Thus, monitoring the boundary of the regulated sector to prevent such excessive recourse is an indispensable step for any future regulation. A direct test will be the treatment of bank credit lines extended to the unregulated financial institutions. To be consistent, the scheme should assign charges on contingent liabilities related to such deals in proportion to the (unregulated) borrowers’ own mismatch. To eliminate arbitrage, such bank credit lines should be treated as uncontingent commitments, and the mismatched asset funding should be fully charged.
The international implementation of such a scheme is certainly complex. Dealing with the liquidity problems of large international banks as well as ensuring a level playing field call for a multi-country jurisdiction. Countries should choose to participate by requiring either all their regulated institutions, or at least the largest ones, to join an international ELIF, pay its liquidity charges, accept its supervision, and count on its support in a systemic crisis.
The establishment of an international ELIF should sort out important commitment problems. Countries that do not join, should not benefit ex post. The scheme should constitute an explicit coordination device for the rescue of large international banks, preventing the issue of burden sharing to be left for difficult ex post negotiations. To the extend that liquidity charges provided a mutually agreed metric for systemic risk and the insurance nature of the scheme, the ELIF would offer a basis for making burden sharing a function of the share of each national banking sector in the liquidity charges paid during the pre-crisis period (rather than politically debatable country quotas).
We argue for mandatory liquidity insurance. Regulated financial institutions would be subject to continuous liquidity charges that would accrue to an Emergency Liquidity Insurance Fund (ELIF) that, in exchange, would provide funding guarantees during systemic crises. The charges would be a function of the degree of maturity mismatch and, possibly, the slope of the term structure.
The aim of the mechanism is to discourage the forms of short term funding that create and amplify propagation risk. It may be also seen as a prepayment of intervention costs, and as a starting step to ensure that liquidity interventions occur on time and are based on ex ante rules.
Brunnermeier, Markus (2009), “Deciphering the Liquidity and Credit Crunch 2007-08,” Journal of Economic Perspectives 23, forthcoming.
Brunnermeier, Markus, Andrew Crockett, Charles Goodhart, Avi Persaud, and Hyun Shin (2009), “The Fundamental Principles of Financial Regulation,” Geneva Reports on the World Economy 11
Calomiris, Charles W., and Charles M. Kahn (1991), “The Role of Demandable Debt in Structuring Optimal Banking Arrangements,” American Economic Review 81, 497-513.
Diamond, Douglas W., and Raghuram G. Rajan (2000), “A Theory of Bank Capital,” Journal of Finance 55, 2431-2465.
Kashyap, Anil K, Raghuram G. Rajan, and Jeremy C. Stein (2008), “Rethinking Capital Regulation,” paper prepared for the Federal Reserve Bank of Kansas City Symposium at Jackson Hole, September.
Suarez, Javier (2008), “The Need for an Emergency bank Debt Insurance Mechanism,” CEPR Policy Insight No. 19, March.
1 An argument for the incentive role of short term liabilities is developed in Calomiris and Kahn (1991) and Diamond and Rajan (2000).